Camera based human fall detection system
Loading...
Date
2014-10
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
UMT Lahore
Abstract
Human fall occurrences are seen as health disasters, especially when elderly people become its victim; as they are more prone to bad health conditions. Detection and notification of a fall, as soon as it occurs can be lifesaving. By detecting falls automatically, as they occur, better timed medical care can be given which can in turn reduce the subsequent medical complications. In this project we proposed an effective fall detection system based on a dataset of videos generated using multiple cameras. We have devised a camera based system for the purpose of detecting falls and indicating them. It detects falls from a video captured from the cameras installed in the suspected area where fall could occur, and then it notifies the concerned parties about the occurrence of the fall. A feature based approach is used that is novel and outperforms other existing approaches on fall detection in terms of accuracy.